New GCC anyone ?

You can never have too many compilers Well, you can, I guess, but the whole point of CODE-RADE is to make any possible combination of applications available to researchers anywhere. So, during the Summer Hackfest it came to our attention that a new version of GCC had been released - GCC-6. It was time to add a few more ticks to the GCC-deploy matrix jobs You can totally have too many compilers Well, enthusiasm is great and all, but it turns out that disk space is limited at tiny build cluster, which made the last few jobs of GCC fail....…

read more...

CODE-RADE : Fermionic Molecular Dynamics

TL;DR : Getting started with Fermionic Molecular Dynamics We’ve recently ported an application from the physical sciences domain used to make nuclear calculations using Fermionic Molecular Dynamics models1. The application was requested2 by Katharine Henninger from iThemba L.A.B.S. Fermionic Molecuar Dynamics in CODE-RADE FMD-deploy was added to CODE-RADE on July 4th. After some discussion3 The application has a pretty simple dependency tree, as seen by the Jenkins config : The dependencies were all passing, so it was just a matter of a few modifications to the source code, the application was included in the CODE-RADE fastrepo. Getting started :tada: Great...…

read more...

Reproducible ASR design

TL;DR : A deeper look at the design of the Reproducible ASR gateway. Background We discussed the motivation and scientific workflows of the Reproducble ASR system in a previous post. Now, we take a bit of a look into what kinds of services are desirable and necessary, as well as how we can design a coherent system. Architecture This science gateway for Automatic Speech Recognition research will be used in a distributed environment and uses many pre-existing services in various service federations, either in the EGI context, the Africa-Arabia Regional Operations Centre, or indeed stand-alone sites. We need the architecture...…

read more...

A plan for the Summer Hackfest

TL;DR : A quick sketch of an environment for reproducible ASR processing We are about to get started on work to develop a virtual research environment for human language researchers, in the context of the Sci-GaIA hackfest. The main focus is on reproducibility and re-analysis of published results. We have thus chosen the most awesome name for the project : RASR - Reproducing Automatic Speech Recognition Background This work comes out of a collaboration with David Risinamhodzi at the University of the North West. That project was in essence a performance study, to investigate the real-world performance of running human...…

read more...